Deviance Information Criterion for Comparing Stochastic Volatility Models

By Andreas Berg, Renate Meyer, Jun Yu. 2004

309 citations recorded by the source corpus.

Bayesian methods have been efficient in estimating parameters of stochastic volatility models for analyzing financial time series. Recent advances made it possible to fit stochastic volatility models of increasing complexity, including covariates, leverage effects, jump components, and heavy-tailed distributions. However, a formal model comparison via Bayes factors remains difficult. The main objective of this article is to demonstrate that model selection is more easily performed using the deviance information criterion (DIC). It combines a Bayesian measure of fit with a measure of model complexity. We illustrate the performance of DIC in discriminating between various different stochastic volatility models using simulated data and daily returns data on the Standard & Poors (S&P) 100 index.

Subfields: Economics and Econometrics, Finance

View Paper: https://www.semanticscholar.org/paper/2b282575a710ef75606ac684ea54487931f3c089

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Economics and Econometrics

In Finance

Citations in the corpus

Estimation methods for stochastic volatility models: a survey (2004, methodology): “Berg et al. (2004) propose to compare models via the information criteria.”
The structure of dynamic correlations in multivariate stochastic volatility models (2009, methodology): “The choice between the CC MSV and DC MSV models can be made using a deviance information criterion (see Berg et al. (2004)).”
Multivariate Stochastic Volatility Models: Bayesian Estimation and Model Comparison (2006, methodology): “1 Since then, BUGS has been employed to estimate univariate SV models in a number of studies (for example, Berg et al., 2004; Lancaster, 2004; Meyer et al., 2003; Selçuk, 2004; Yu, 2005). Furthermore, Berg et al. (2004) showed that model selection of alternative univariate SV models is easily perfor”
The dynamics of commodity prices (2013, unknown): “(2002) and in particular applied to stochastic volatility models by Berg et al. (2004). The DIC can be regarded as a generalization of the Akaike Information Criterion (AIC), and trades off adequacy and complexity in a”
Multivariate stochastic volatility (2009, methodology): “Recently, Berg et al. (2004) showed that model selection of alternative univariate SV models can be performed easily using the deviance information criterion (DIC) proposed by Spiegelhalter et al. (2002), while Yu and Meyer (2004) compared alternative MSV models using DIC.”
Multivariate Stochastic Volatility: A Review (2006, methodology): “Recently, Berg et al. (2004) showed that model selection of alternative univariate SV models can be performed easily using the deviance information criterion (DIC) proposed by Spiegelhalter et al.”
Exchange Rates and Oil Prices: A Multivariate Stochastic Volatility Analysis (2012, unknown): “Benassy-Querea, Mignonb, and Penot (2007), Coudert, Mignon, and Penot (2007), Chen and Chen (2007), Ayadi (2005), Chaudhuri and Daniel (1998) and Krugman (1984) all provide evidence supporting this view.”
Bayesian Model Uncertainty In Smooth Transition Autoregressions (2006, methodology): “For successful implementations of DIC we refer to Zhu and Carlin (2000) (spatio-temporal hierarchical models) and Berg et al. (2004) (stochastic volatility models), to name a few.”
Model Selection Methods for Mixture Dichotomous IRT Models (2009, background): “Berg et al. also reported results for DIC consistent with other results for nonmixture model comparisons. Richardson (2002) provided an example of the performance of DIC for mixture models and pointed out that pD apparently provides a sufficient penalty only for the simplest cases.”
Variational approximations in Bayesian model selection for finite mixture distributions (2007, methodology): “attractive option for modern applications; for instance, Berg et al. (2004) apply the DIC to”
Volatility dynamics for the S&P 500: Further evidence from non-affine, multi-factor jump diffusions (2012, result): “For SV models with jumps the DIC has been studied in detail by Berg et al. (2004), who find that the ranking of models generated by the DIC (a lower value being indicative of a better model) is comparable to the ranking generated by Bayes factors.”
On Leverage in a Stochastic Volatility Model (2005, unknown): “For example, Melino and Turnbull (1990) show that prices of European call options on currencies based on the basic SV models are more accurate than those based on the Black–Scholes model. Kim et al. (1998) provide evidence of better in-sample-fit of the basic SV model relative to GARCH-type models. ”
Performance of deviance information criterion model selection in statistical catch-at-age analysis (2008, methodology): “Bayesian theory would suggest averaging models based upon Bayes factors, but there are conceptual and practical difficulties, including difficulties in obtaining reliable estimates of Bayes factors from MCMC samples (Kass and Raftery, 1995) and concerns that Bayes factors may not consider the approp”